dd-audit-security-investigation
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and interpret Datadog audit logs, which creates a surface where malicious entries in the logs could attempt to misdirect the agent.\n
- Ingestion points: Audit log data is retrieved through pup audit-logs search and processed by the agent to answer security questions in SKILL.md.\n
- Boundary markers: There are no explicit instructions or markers used to delimit retrieved log data from the agent's internal reasoning or instructions.\n
- Capability inventory: The skill relies on shell command execution for pup (network queries) and jq (data processing).\n
- Sanitization: While jq is used to filter fields, the content within those fields is not sanitized for potential prompt injection patterns.\n- [COMMAND_EXECUTION]: The skill functions by directing the agent to execute shell commands using the pup and jq utilities.\n
- Evidence: Multiple shell code blocks in SKILL.md demonstrate the use of pup audit-logs search and pup auth login commands to interact with external APIs.
Audit Metadata